一般口演（Oral）

Oral Connectome 一般口演コネクトーム

7月25日（木）15:40～15:55　第8会場（朱鷺メッセ　3F　303+304） 1O-08a2-1 Fully automated data processing for mapping connectivity of the marmoset prefrontal cortex Henrik Skibbe（Skibbe Henrik）¹，Akiya Watakabe（Watakabe Akiya）²，Ken Nakae（Nakae Ken）¹，Carlos Enrique Gutierrez（Gutierrez Carlos Enrique）³，Alexander Woodward（Woodward Alexander）⁴，Hiromichi Tsukada（Tsukada Hiromichi）³，Rui Gong（Gong Rui）⁴，Junichi Hata（Hata Junichi）^5,6，Kenji Doya（Doya Kenji）³，Hideyuki Okano（Okano Hideyuki）^5,6，Tetsuo Yamamori（Yamamori Tetsuo）²，Shin Ishii（Ishii Shin）¹ ¹Department of Systems Science, Kyoto University, Kyoto, Japan ²Laboratory for Molecular Analysis of Higher Brain Function, RIKEN Center for Brain Science, Wako, Japan ³Neural Computation Unit, Okinawa Institute of Science and Technology, Okinawa, Japan ⁴Neuroinformatics Unit, RIKEN Center for Brain Science, Wako, Japan ⁵Laboratory for Marmoset Neural Architecture, RIKEN Center for Brain Science, Wako, Japan ⁶Department of Physiology, Keio University School of Medicine, Tokyo, Japan The Japan Brain/MINDS project [1,2] is conducting a large scale connectivity study on marmoset brains. In the project, two-photon microscopy fluorescence images of axonal projections are used to collect the neuron connectivity from defined brain regions at the mesoscopic scale. The processing of the images requires the detection, segmentation, and registration of the axonal tracer signal. We have developed a fully automated processing pipeline that processes and maps multi-modal brain imaging data, such as tracer, Nissl, and MRI data, to a common brain image space before connectivity calculations are being carried out. The pipeline incorporates state-of-the-art image processing and machine learning techniques to extract and map all relevant information in a robust manner. [1] BrainMinds http://brainminds.jp/en/ [2] H. Okano et al., ""Brain/minds: A Japanese national brain project for marmoset neuroscience"";, Neuron, 2016. This research is supported by the program for Brain Mapping by Integrated Neurotechnologies for Disease Studies (Brain/MINDS) from Japan Agency for Medical Research and development, AMED		7月25日（木）15:55～16:10　第8会場（朱鷺メッセ　3F　303+304） 1O-08a2-2 マカクサル30頭における高解像度マルチモーダル画像による脳コネクトーム(MacCP30) Takuya Hayashi（林拓也）¹，Joonas Autio（Autio Joonas）¹，Takayuki Ose（合瀬恭幸）¹，Masahiro Ohno（大野正裕）¹，Kantaro Nishigori（西郡寛太郎）¹，Akihiro Kawasaki（川崎章弘）¹，Chiho Takeda（武田千穂）¹，Chihiro Yokoyama（横山ちひろ）¹，Timothy Coalson（Coalson Timothy）²，Chad Donahue（Donahue Chad）²，Stephen Smith（Smith Stephen）³，David Van Essen（Van Essen David）²，Matthew Glasser（Glasser Matthew）² ¹理研BDR　脳コネクトミクスイメージング ²Dept Neurosci, Washington Univ Med Sch, St Louis ³FMRB, Wellcome Center for Integ Neuroimaging, Univ Oxford, UK Studies of non-human primates provide invaluable information relevant to understanding how the human brain evolved its unique structure, function and connectivity. Recent studies of the human brain using high quality and quantity, non-invasive and multi-modal MRI from the HCP have enabled in-depth characterization of brain organization including a new cortical parcellation (Glasser, 2016), functional connectome (Smith, 2015). Here we introduce a macaque functional connectome generated from high-resolution multi-modal MRI data from thirty macaques, MacCP 30. MacCP 30 is based on brain MRI data in macaque monkeys (M. fascicularis: N=14, age= 4.9 ± 2.1 years; and M. mulatta: N=16, age =5.2 ± 1.7). All animals were scanned under anesthesia, and one animal was also scanned with a contrast agent, MION in alert. T1w and T2w structural images (0.5mm isotropic voxels) and resting-state fMRI (1.25mm isotropic voxels, TR=0.75 sec) were obtained using a 3T MRI scanner (MAGNETOME Prisma, Siemens, Erlangen, Germany) and 24-ch multi-array RF coil (Takashima Seisakusho KK, Hino, Japan). Structural images were preprocessed using a non-human primate version of the HCP pipeline (Donahue, 2016; Glasser, 2013) and fMRI data were denoised with spatial independent component analysis (ICA) and Wishart filtering (Glasser, 2018). Functional connectome was analyzed for gradients (Glasser, 2016), classified into two modes by k-means clustering, and compared between species and anesthetic levels (light vs deep; awake vs anesthetized). Macaque cortical myelin, thickness, sulcal depth, and dense functional connectivity showed distinct cortical patterns. The connectivity gradient was high along the lateral, ventral, and medial prefrontal-premotor border, near area 12, and along the superior temporal and cingulate gyri, which colocalized with borders between the two functional modes. The gradient ridges generally showed lower correlation of dense connectome across species and anesthesia levels. These finding suggest that MacCP30 dataset demonstrates a complex and dense organization of functional connectome. Two functional modes distinguished sensory-motor and association-perception areas like in humans (Smith, 2015; Vidaurre, 2017), and the transition zone between these modes varied in connectivity pattern depending on species and consciousness level, suggesting that functional gradients may underlie cortical dynamics and evolution.

7月25日（木）16:10～16:25　第8会場（朱鷺メッセ　3F　303+304） 1O-08a2-3 Subcortico-centric view of macaque neocortex investigated using resting-state functional MRI Joonas A Autio（Autio Joonas A）¹，Matthew F Glasser（Glasser Matthew）^2,3，Atsushi Yoshida（Yoshida Atsushi）¹，Takayuki Ose（Ose Takayuki）¹，Kantaro Nishigori（Nishigori Kantaro）⁴，Masataka Yamaguchi（Yamaguchi Masataka）¹，Chihiro Yokohama（Yokohama Chihiro）¹，Masahiro Ohno（Ohno Masahiro）¹，David C Van Essen（Van Essen David C）²，Takuya Hayashi（Hayashi Takuya）¹ ¹Laboratory for Brain Connectomics Imaging, RIKEN Center for Biosystems Dynamics Research, Kobe, Japan ²Department of Neuroscience, Washington University in St. Louis, Missouri, USA ³Department of Radiology, Washington University in St. Louis, Missouri, USA ⁴Sumitomo Dainippon Pharma Co., Ltd, Osaka, Japan An important aspect to understand evolutionary differences across primate species may be through conserved subcortical circuitry and diversification of neocortical inputs. Here, we demonstrate neocortical profiles of major subcortical structures using Human Connectome Project (HCP)-style resting-state functional MRI (rfMRI) connectivity (FC) in alert (N=2) and anesthetized (N=30) macaque monkeys. The experiments were performed in a clinical 3 T MRI scanner (MAGNETOM Prisma, Siemens, Erlangen, Germany) equipped with 80 mT/m gradients in combination with a custom-made 24-channel macaque coil. We used `HCP-style' data acquisition (structural and functional) and analysis pipelines customized for the macaque. Our results reveal that the major subcortical 'limbic and associative' structures (amygdala, accumbens, hippocampus and caudate nucleus) have largely overlapping neocortical FC profiles. The predominantly somatomotor community (putamen, globus pallidus, cerebellum and brain stem) also shows a high degree of spatial overlap but these are strongly anticorrelated with associative neocortical connectivity profiles. This dichotomy is supported by a fuzzy (soft) C-means clustering using all brain greyordinate edges (n=26,020, edges=2.7× 109). We also examined the association between neocortical myelination and subcortico-neocortical FC, as cortical myelin content provides a useful indicator of transitions between associative and somatomotor regions (Glasser et al., 2013, Donahue et al., 2018). Indeed, the neocortical myelination was highly anticorrelated with FC from associative subcortical structures (i.e. amygdala R=-0.73, accumbens R=-0.71, hippocampus R=-0.37 and caudate nucleus R=-0.40, p<10-6), whereas no significant association was found with FC from cerebellum and putamen. Taken together, these findings may provide a valuable clue to subcortico-centric view on the evolution of primate brain.		7月25日（木）16:25～16:40　第8会場（朱鷺メッセ　3F　303+304） 1O-08a2-4 マイクロコネクトーム上での興奮性ー抑制性の情報流 Motoki Kajiwara（梶原基）¹，Felix Goetz（ゲーツフェリックス）²，Akitoshi Seiyama（精山明敏）¹，Masanori Shimono（下野昌宣）¹ ¹京都大学医学部人間健康科学科 ²National Central University, Taoyuan, Taiwan Our brain forms a nonuniform, multi-scale network of neurons. In this network the neuronal assemblies generate spike patterns, which are electric signals at millisecond time scale that regulate the cognitive functions. Research on these complex networks between neurons started more than 150 years ago, when Santiago Ramon y Cajal observed the structure of neurons using staining techniques. After his era various recording technologies have been developed in order to observe the collective activity of neuronal assemblies. The technologies have enabled us to record electric signals simultaneously from huge numbers of neurons. Subsequently, techniques were developed to reconstruct the interaction networks between huge number of neurons from their spiking activities. This allowed to elucidate the Microconnectome and the comprehensive and quantitative characteristics of interaction networks among neurons (Shimono, Beggs, 2014). Here, we measured electrical activities from in vitro brains slices of mice with a MEA (multi-electrode array) to analyze the network architectures of cortical neurons. We recently introduced a slit light irradiation type 3D scanner for position identification of brain slices in whole brain MRI (Ide et al., Under Review). These experimental setup enabled us to measure thousands of neurons simultaneously, which allowed us to perform a detailed analyses on the neuronal assemblies.The analyses mainly consists of two steps: First, Transfer entropy was adopted from previous research to reconstruct the neural network. Briefly, Transfer Entropy quantifies the amount of information transferred between neurons and is suitable for the effective connectivity analysis of neural networks. In addition our study distinguishes between excitatory synapses and inhibitory synapses using a newly developed method called sorted local transfer entropy (Goetze et al., 2015). Second, we also applied methods from graph theory to evaluate the network architecture. Graph theory has been applied to various networks from complex systems, studied in sociology and biology. In this study, by applying the new sorted local transfer entropy method to the recorded data, we focus on the cell categories of excitatory and inhibitory cells. In the conference, we will report our findings about the neural networks in the cortical region using the new method.